From the 50 Cent Party to Spamouflage Dragon: how China conducts online propaganda wars

At the onset of 2010’s, massive protests shook the Northern African and Middle East regions like it hadn’t occurred in a long time. In what would be later dubbed Arab Spring, autocrats trembled at the sight of ferocious citizens demanding decent standards of living and a rupture from the miseries of the past. From Tunisia to Egypt, from Syria to Bahrein, the youth took on the streets to oust long-lived regimes; and it did so by making intensive use of social media.

The Communist Party of China (CPC) duly took note: what materialized in the Arab world had to be avoided at all costs in the former celestial empire. In 2013, after being recently appointed as President of the People’s Republic of China (PRC), Xi Jinping spelled out his intention to “retake the Internet battlefield.” Meanwhile, an article from the Party-owned newspaper Beijing Daily “stressed the need to make the internet the main battlefield of public opinion struggles”. The message was clear: positive views of the Party and of its ideological underpinnings (Marxism and socialism with Chinese characteristics) had to dominate the virtual landscape, or else the regime would have risked losing ground to rebellious voices and popular discontent.

That’s when the 50 Cent Party increased the scale of its operations. The 50 Cent Party (or 50 Cent Army) is a group of government employees or ad hoc hired citizens who pose as regular online users and flood the Chinese internet with pro-regime comments. There are no precise estimates on the number of members, but these range from hundreds of thousands to 2 million.

While the group first surfaced in 2004, more information was collected following a 2014 data leak involving thousands of emails. The term 50 Cent refers to the alleged sum paid to the employees for every comment posted, but no evidence was found to substantiate this claim.

A Harvard study from 2017 sheds light on the figures, objectives and tactics of the group. Drawing from empirical data, researchers found that approximately 448 million comments posted on the Chinese internet in 2013 originated from government paid commenters. Around 52% of the comments appeared on Party-affiliated websites, while the rest was scattered through social media platforms. The study also found high levels of coordination by the 50 Centers, both in the timing and the type of content to address, which led authors to speculate that “these efforts may be directed from the highest levels of the regime”.

The tactics used by the 50 Cent Party seemed to follow two basic principles: 1) do not engage on controversial issues and 2) stop discussions with collective action potential. Indeed, contrary to what previously claimed, most 50 Centers did not argue with critics of the government but changed the subjects of their comments to distract the public from particularly sensitive issues. In short, online discussions which could have potentially led to mass protests were to be inundated by positive messaging. The authors of the study divided the examined comments into several categories and found that most of them (80%) consisted of cheerleading for China, while only a handful engaged in arguments or taunted western countries and values. One example of cheerleading comment reads: “Respect to all the people who have greatly contributed to the prosperity and success of the Chinese civilization! The heroes of the people are immortal.”

While this research offers quite outdated figures, it gives us a glimpse of the huge propaganda machine at work behind the Great Firewall. In the present days, the scale of operations might actually have increased: following the COVID-19 outbreak in Wuhan, the Chinese regime had to batten down the hatches and wash its stained reputation. For that reason, the Communist Party engaged in a not-so-covert online propaganda campaign directed at international audiences, including the Chinese diaspora, which also involved discrediting the United States and spreading conspiracy theories. The multitude of Chinese fake accounts found operating on social media certainly signalled a more assertive online attitude by the regime, but it remains difficult to ascertain what role the 50 Cent Party played, if any.

This leads us to Spamouflage Dragon, a Chinese spam network emerged over the last few years. The social media analysis company Graphika first spotted the network in 2018 and baptized it Spamouflage as the fake and repurposed accounts making up the network camouflaged their political messaging with innocent content (showing, for instance, cute animals and dancing girls).

Spamouflage mostly operates on Facebook, Twitter and YouTube. In August 2019, all three companies exposed a cross-platform operation and took down several accounts affiliated with the network. It had mainly two targets: pro-Hong Kong protesters and Guo Wengui, a fugitive Chinese billionaire with ties to the Trump administration and several criminal charges pending against him.

The covid-19 pandemic changed the network’s priorities and Spamouflage began spreading positive messaging on the regime’s reaction to the outbreak.

As the pandemic worsened, relations between China and the United States deteriorated; in June 2020, Spamouflage surfaced on the US internet with accounts employing AI-generated profile pictures and using English language. The accounts exploited the public tensions erupted after the killing of George Floyd to foment the country’s racial divisions and also used Trump’s invectives against TikTok to attack his administration. The European Union as well was targeted by virus-related Chinese propaganda, but the involvement of Spamouflage is still unclear.

Ultimately, while it is hard to evaluate whether online campaigns conducted by the 50 Cent Party or Spamouflage reached their objectives, the potential behind the regime’s online propaganda efforts is becoming all the more evident. Spamouflage’s information warfare was deemed clumsy and poorly orchestrated and not everyone agrees that the Chinese regime won back its reputation. Yet, the propagation of inflammatory content and manipulative messages polarizes society and pollutes our democracy. Cutting-edge technologies like the new AI language model GPT-3 could give disinformation a massive boost by providing malicious actors with the capabilities to produce vast amounts of disinformation indistinguishable from human-generated texts.

Collective action and future-proof measures are urgently needed to safeguard online public spaces.